Prediction of steroid-sparing agent use in childhood idiopathic nephrotic syndrome

Abstract

About half of children with steroid-sensitive idiopathic nephrotic syndrome (INS) will develop steroid dependency or a frequently relapsing course requiring steroid-sparing agents (SSA). Because of the adverse effects of prolonged steroid treatment, the early identification of children at high risk of requiring SSA may be a useful diagnostic tool to tailor the therapeutic strategy. The aim of this study was to identify predictors of the need for SSA and derive a predictive model. This was a retrospective hospital-based cohort study which included all children with steroid-responsive INS followed for at least 4.5 months. Cox regression modeling and decision curve analysis were performed. A total of 120 children (81 boys) with INS were included and followed up for a median time of 6.7 (range 0.4-24.1) years. Median age at diagnosis was 3.4 years. Seventy-two (60 %) children required a SSA after a median time of 10 months following initial diagnosis. Male children, age at disease onset, methylprednisolone pulse use, and time to achieve first remission were significantly associated with the outcome. Time to achieve remission only remained significant after adjustment: hazard ratio (HR) =1.9 [95 % confidence interval (CI) 1.5-2.5] if considered as a continuous variable, and HR=4.1 (95 % CI 1.9-8.6) when dichotomized using the 9-day threshold. The area under the receiver operating curve of the related predictive model was 0.81 (95 % CI 0.74-0.89), and the decision curve analysis demonstrated that this model performed better than any other strategy. Time to first remission is a strong predictor of the need for SSA in pediatric INS. Further prospective and impact studies are warranted to confirm the accuracy and benefit of our prediction model.